The ATP Level in the mPFC Mediates the Antidepressant Effect of Calorie Restriction

Abstract

Food deprivation can rescue obesity and overweight-induced mood disorders, and promote mood performance in normal subjects. Animal studies and clinical research have revealed the antidepressant-like effect of calorie restriction, but little is known about the mechanism of calorie restriction-induced mood modification. Previous studies have found that astrocytes modulate depressive-like behaviors. Inositol 1,4,5-trisphosphate receptor type 2 (IP3R2) is the predominant isoform in mediating astrocyte Ca²⁺ signals and its genetic knockout mice are widely used to study astrocyte function in vivo. In this study, we showed that deletion of IP3R2 blocked the antidepressant-like effect induced by calorie restriction. In vivo microdialysis experiments demonstrated that calorie restriction induced an increase in ATP level in the medial prefrontal cortex (mPFC) in naïve mice but this effect disappeared in IP3R2-knockout mice, suggesting a role of astrocytic ATP in the calorie restriction-induced antidepressant effect. Further experiments showed that systemic administration and local infusion of ATP into the mPFC induced an antidepressant effect, whereas decreasing ATP by Apyrase in the mPFC blocked calorie restriction-induced antidepressant regulation. Together, these findings support a role for astrocytic ATP in the antidepressant-like effect caused by calorie restriction.

Keywords: ATP; Astrocyte; Calorie restriction; Depression; IP3R2.

Fig. 1

Calorie restriction (CR) is sufficient to induce an antidepressant-like effect in adult C57BL/6J mice. A Weight changes for 10 consecutive days of CR and ad lib mice (F_1,18 = 181.2, P <0.0001, n = 10). B The time mice remained on the rotarod in the rotarod test. C–D The total distance (C) and the numbers of vertical rearing movements (D) for ad lib and CR mice in the open field test (C: t₁₈ = 0.7683, P = 0.4523, n =10; D: t₁₈ = 0.9998, P = 0.3307, n = 10). E The center area time for ad lib and CR mice in the open field test (t₁₈ = 2.212, P = 0.0401, n = 10). F The time on the open arms in the elevated plus maze test (t₁₇ = 3.944, P = 0.001, n_{ad lib} = 10, n_CR = 9). G The total duration of immobility in the forced swimming test (t₁₈ = 5.425, P <0.0001, n = 10). ad lib: ad libitum; CR: calorie restriction; The data are presented as the mean ± SEM (repeated measures ANOVA (A–B); two-tailed unpaired t test (C–G); *P <0.05, **P <0.01, ****P <0.0001; comparisons with no asterisk had a P >0.05 and were considered not significant).

Fig. 2

IP3R2 knockout blocks the antidepressant-like effect induced by calorie restriction (CR). A Weight changes of IP3R2-WT and -KO mice after ad lib or CR (F_3,36 = 22.42, P < .001, n = 10). B The time IP3R2-KO mice remained on the rotarod in the rotarod test (F_3,36 = 0.1407, P = 0.0.9350, n = 10). C–D The total distance (C) and the numbers of vertical rearing movements (D) in the open field test (C: F_3,36 = 0.7393, P = 0.4206, n = 10; D: F_3,36 = 0.5193, P = 0.6649, n = 10). E The center area time in the open field test (F_3,36 = 8.527, P = 0.0222, n = 10; WT_{ad lib} vs WT_CR, P = 0.0214). F The time on open arms in elevated plus maze (F_3,36 = 8.527, P = 0.0062, n = 10; WT_{ad lib} vs WT_CR, P = 0.0125). G The total duration of immobility in the forced swimming test (F_3,36 = 6.531, P = 0.0012, n = 10; WT_{ad lib} vs WT_CR, P = 0.0107; WT_{ad lib} vs KO_{ad lib,} P = 0.0317; KO_{ad lib} vs KO_CR, P = 0.4118). WT: wildtype mice; KO: IP3R2-knockout mice. The data are presented as the mean ± SEM (*P <0.05; repeated measures ANOVA (A–B); one-way ANOVA with Fisher's LSD multiple comparisons post hoc test (C–G); comparisons with no asterisk had a P >0.05 and were considered not significant).

Fig. 3

The ATP level is increased by calorie restriction (CR) but this effect is absent in IP3R2-KO mice. A In vivo microdialysis assays showing the extracellular ATP levels in the mPFC, hippocampus, and amygdala following ad lib and CR treatment (t₁₂ = 6.224, P <0.0001, n = 7; t₁₂ = 3.879, P = 0.0022, n = 7; t₁₂ = 4.445, P = 0.0008, n = 7). B In vivo microdialysis assays showing the extracellular ATP levels in the mPFC, hippocampus, and amygdala in IP3R2-KO mice (t₁₂ = 6.224, P <0.0001, n = 7; t₁₂ = 3.879, P = 0.0022, n = 7; t₁₂ = 4.445, P = 0.0008, n = 7). C The ATP levels in the culture medium of astrocytes from IP3R2-KO mice compared with WT mice (t₁₂ = 11.26, P <0.0001, n = 7). D The ATP levels in the culture medium of neurons from IP3R2-KO and WT mice (t₁₂ = 0.3719, P = 0.7164, n = 7). E In vivo microdialysis assays showing extracellular ATP levels in the mPFC of IP3R2-KO mice that received ad lib and CR treatment (t₁₄ = 0.3115, P = 0.7600, n = 8). The data are presented as the mean ± SEM [two-tailed unpaired t test (A–E); **P <0.01, ***P <0.001, ****P <0.0001; comparisons with no asterisk had a P >0.05 and were considered not significant].

Fig. 4

Behavioral changes following the modulation of systemic ATP levels. A Quantification of the total duration of immobility as a percentage of the vehicle after i.p. infusion of ATP (125 mg/kg) in C57BL/6J mice (t₁₈ = 6.407, P <0.0001, n = 10). B As in A, but for the open field test (t₁₈ = 0.3533, P = 0.7279, n = 10). C Quantification of the total duration of immobility as a percentage of the vehicle after i.p. infusion of ATP (125 mg/kg) in IP3R2-KO mice (F_3,36 = 8.527, P = 0.0062, n = 10; WT_Saline vs WT_ATP, P = 0.0008; WT_Saline vs KO_Saline, P = 0.0003; KO_Saline vs KO_ATP, P = 0.0001) . D As in C, but for the open field test (F_3,36 = 1.161, P = 0.3381, n = 10). The data are presented as the mean ± SEM [two-tailed unpaired t test (A, B); one-way ANOVA with Fisher's LSD multiple comparison post hoc test (C, D); ***P <0.001, ****P <0.0001; comparisons with no asterisk had a P >0.05 and were considered not significant].

Fig. 5

Behavioral changes following the modulation of mPFC ATP levels. A The injection site in the mPFC (scale bar, 500 μm). B Quantification of the total duration of immobility as a percentage of the vehicle after mPFC infusion of ATP (25 μmol/L) in C57BL/6J mice (t₁₈ = 4.624, P = 0.0002, n = 10). C As in B, but for the open field test (t₁₈ = 0.0051, P = 0.9959, n = 10). D Quantification of the total duration of immobility as a percentage of the vehicle after mPFC infusion of the hydrolase of ATP Apyrase (300 U/ml) in C57BL/6J mice (F_3,36 = 5.948, P = 0.0025, n = 10; Con_ACSF vs CR_APY, P = 0.0498; Con_ACSF vs CR_ACSF, P = 0.0374; CR_ACSF vs CR_APY, P = 0.0115). E As in F, but for the open field test (F_3,36 = 0.09148, P = 0.9643, n = 10). The data are presented as the mean ± SEM [two-tailed unpaired t test (B, C); one-way ANOVA with Fisher's LSD multiple comparison post hoc test (D, E); *P <0.05, ***P <0.001; comparisons with no asterisk had a P >0.05 and were considered not significant].

Fig. 6

Reduced sEPSC frequency in the mPFC in calorie-restricted (CR) mice. A sEPSC recordings in mPFC pyramidal neurons from ad lib and CR mouse slices (scale bars, 20 pA, 2 s). B–C Bar graphs and cumulative plots showing sEPSC frequency and amplitude in ad lib and CR mouse slices (B: t₁₂ = 2.649, P = 0.0212; C: t₁₂ =0.409, P = 0.6898; n_{ad lib} = 8, n_CR = 6). D sIPSC recordings in mPFC pyramidal neurons from ad lib and CR mouse slices (scale bars, 20 pA, 2 s). E–F Bar graphs and cumulative plots showing sIPSC frequency and amplitude in ad lib and CR mouse slices (E: t₁₅ = 5.386, P = 0.5981; F: t₁₈ = 1.271, P = 0.2231; n_{ad lib} = 9, n_CR = 8). ad lib, ad libitum; CR, calorie restriction. The data are presented as the mean ± SEM [two-tailed unpaired t test (A–J); *P <0.05; comparisons with no asterisk had a P >0.05 and were considered not significant].